difference between multiexon spanning and tandem cassette exon

[Alex PG]

Hi all,

Thank you again for all your help so far.

I wanted to ask what the exact difference is between a multiexon spanning event and a tandem cassette exon. How exactly are these events calculated?

In my cancer vs. normal comparison analysis, I identified 101 multiexon spanning events and 8 tandem cassette exon events. I am trying to better understand the distinction between them, as it is not entirely obvious to me from the Voila plots.

Thank you so much in advance.

All the best,
Alex

[San Jewell]

Hi Alex!

Thank you for asking here. I will use the same images from the docs at https://biociphers.bitbucket.io/majiq-docs/modulizer/event-types.html#multi-exon-spanning to explain. 

Multiexon spanning is a very simple pattern match which only checks whether there are one or more junctions which happen to splice together exons which would splice out more than one other exon. For example, in the example, the 21 read junction splices out exons 23, 24, and 25. And the 19 read junction splices out exon 24 and 25. This event type isn't mutually exclusive with most others and will often appear alongside other event types, including tandem cassette exons. 

Tandem cassette exons are a stricter type of pattern match which looks for one long spanning junction connecting two flanking exons (in this example exons 4 and 7), which must also be connected by some series of inner exons in a constitutive manner. The event is defined by a choice between one 'skip' (5 read junction) OR a series of constitutive exons greater than one (22, 24, 28 read junctions). This is by definition also multiexon spanning. The multiexon spanning example above is not a tandem cassette because exons 24 and 25 are not spliced together by any junction. 

Let me know if this helps to clarify!

Thank you, 

-San